A multimodal intervention reduces computed tomography use in blunt head injury

Computed tomography (CT) is frequently used in emergency departments (EDs) for the evaluation of blunt head injury. Unnecessary CT utilization leads to excessive cost, radiation exposure, and operational inefficiencies. Despite validated clinical decision rules, head CT utilization continues to rise, and the need remains for effective implementation strategies.

We performed a retrospective pre-post analysis after implementing a multi-modal intervention at three EDs within an academic health system. The intervention consisted of a guideline supporting a choice between the NEXUS Head CT Rule or Canadian CT Head Rule, educational sessions, monthly provider feedback, and an EHR-integrated clinical decision support (CDS) tool. All ED encounters among patients ages 18-64 were included after excluding those with high-risk features. Data were analyzed with logistic regression with generalized estimating equations and XmR statistical process control charts. A subgroup analysis was restricted to patients with traumatic chief complaints. The primary outcome was the rate of CT brain utilization. Secondary outcomes included rate of CT studies positive for radiographic traumatic brain injury (TBI), 72-h ED return visits, and 72-h ED return visits with positive head CTs.

Logistic regression showed a significant associated reduction in odds of CT brain studies among the pooled sites in the post-intervention period for all encounters (Adjusted OR 0.91, 95% CI 0.89-0.94) and for traumatic chief complaints (Adjusted OR 0.89, 95% CI 0.85-0.94). SPC analysis showed special cause variation both immediately after the intervention and throughout the post-intervention year. There was no difference in the rate of positive head CT post-intervention (OR 0.92, 95% CI 0.72-1.16), rate of 72-h returns (OR 0.96, 95% CI 0.82-1.12), or 72-h returns with positive radiographic TBI (OR 0.78, 95% CI 0.22-2.82).

Deploying an institutional guideline supported by educational sessions, provider feedback, and integrated CDS was associated with a reduction in CT head studies.